The A Hybrid Weighted Ensemble classifier model for medical databases

Abstract

Extreme learning approaches are now widely used to identify and diagnose medical conditions
for large databases. Ensemble classifier was a key study model for extreme learning machines
for real-time applications due to its great performance and processing speed. Due to the static
weight selection of the output layer hidden, standard extreme learning methods are unable to
estimate the error rate. This research introduces a novel weighted extreme learning machine
(WELM) for medical condition prediction. The basic goal of the weighted extreme learner is
to define high-dimensional data for illness prediction. Typically, the proposed ensemble model
is created and deployed to improve cancer prediction using high-dimensional data. Using
several ensemble learning models such as random forest, neural networks, ACO+NN, and
PSO+NN, we evaluated the performance of the WELM model suggested in this paper. Test
outcomes are examined in a variety of medical datasets, including liver, diabetes, ovarian, and
DLBCL-Stanford. The WELM presented is highly computationally efficient in terms of true
positive rate, error rate, and accuracy, according to the results.